Showing papers on "Overhead (computing) published in 1975"

Design of self-checking software

Abstract: This paper discusses different techniques for constructing a piece of self-checking software for systems where ultra-reliability is required. Self-checking software can be designed to detect software errors, to locate and to stop the propagation of software errors, to assist in the recovery from errors and to verify the integrity of the system. Self-checking techniques can be implemented in the program to check the function, the control sequence and the data of a process. The functional aspects of a process can be monitored to detect infinite loops, incorrect loop terminations, illegal branches and wrong branches. The validity of data of a process can be assured by performing checks on the integrity of data values, the integrity of data structures and the nature of data values. These self-checking capabilities should be implemented during the initial stage of program development. The cost-effectiveness of each technique in the particular operating environment should be evaluated. Only the most cost-effective techniques should be retained and overhead can be reduced considerably by implementing these techniques in hardware.

Switching system for non-symmetrical sharing of computer peripheral equipment

Abstract: Switching apparatus and methods are disclosed by which one or more secondary computers of a loosely-coupled multiprocessing system may communicate directly with selected items of active peripheral equipment of a main computer of the system. This invention, which is an improvement over device-sharing means and indirect-accessing methods of the prior art, may be applied to existing computer systems without modification to any element of their hardware and without modification to the operating system software of the main computer. An essentially autonomous "intelligent switch," connected to an input/output interface (bus) of each processor and completely transparent to the main processor, is used to switch one or more items of operating peripheral equipment from the main computer to a secondary computer in response to asynchronous service requests from the secondary computer. Two modes of operation are provided: Interface Capture and Data Capture. Interface Capture provides a means for temporarily borrowing items of peripheral equipment having off-line operating capabilities, such as buffered line-printers, and is performed during a normal off-line operation of such equipment. With Interface Capture, the requesting secondary computer obtains complete control of the captured item of peripheral equipment for an arbitrary period of time. Data Capture, on the other hand, is designed for use primarily with direct-access auxiliary storage devices and is performed while a device is executing a read or write operation initiated by the main computer. Data Capture provides a secondary computer with the means for directly accessing shared data space in auxiliary storage without, however, requiring the secondary computer to duplicate or replicate the access-method software and input/output hardware capabilities of the main computer system. Instead, the secondary computer draws parasitically upon the hardware and software resources of the main computer to access the requested data space. Unlike indirect-accessing methods of the prior art, however, this is accomplished without burdening the main processor, its main storage, and its input/output channels with the overhead tasks of buffering data blocks and dispatching data to or from the requesting secondary computer.

Blow-molded articulated overhead door

Abstract: An articulated overhead door constructed of blow-molded material sections for use on building structures such as industrial, commercial, and residential garages.

Ceiling system and prefabricated overhead building assembly using this system

Abstract: A prefabricated overhead building assembly and components making up this assembly are disclosed herein. The assembly includes a prefabricated overhead building structure, for example, a prefabricated roof, and a suspended ceiling system connected to and directly under the overhead structure. This ceiling system is comprised of a number of support runners and connectors which connect the support runners together to form a ceiling support grid supporting, for example, ceiling tile. The ceiling system also includes arrangements of cooperating components which fixedly connect the support grid to the overhead structure at alternative spaced positions below the overhead structure.

Effective Comparison of Unconstrained Optimization Techniques

Abstract: The paper presents a means of attempting to account for overhead, as well as function evaluations, in evaluating unconstrained optimization techniques. Criteria are established which compare techniques on classes of algorithms. While not completely machine independent, and certainly not programmer independent, the new method eliminates much of the machine dependency of earlier criteria. The criteria are applied to three known and one new quasi-Newton algorithm, with interesting results.

Safety stop for overhead track switch

Abstract: A safety stop for overhead track switches prevents a trolley on the track from rolling off an open end if the switch malfunctions. An additional feature of the safety stop are beveled portions thereof which help align the tracks being switched with the stationary portion of the track.

Machine-inspired enhancements of the simplex algorithm

Abstract: Using spare storage and over-lapping overhead available during the generation of the pricing form in the revised-simplex algorithm allows the production of additional reduced costs of variables which may be effectively used to avoid redundant pivots and to predict actual gains. Results from a simulation indicate a dramatic improvement over all other meaningful selection methods.

Bounds on the speed-up of parallel evaluation of recurrences

Abstract: : To understand the performance of parallel computers such as ILLIAC IV and C.mmp, the largest speed-up that can be obtained for a given task must be known. If there are k processors, the largest speed-up that can be achieved is k and the authors call this optimal speed-up. The speed-ups in general depend on the parallel decompositon of a particular computing task and the various aspects of the multiprocessing system, including memory contention, process communication, operating system overhead, etc. In this paper, the authors concentrate on the issue of decomposing tasks, and assume that the multiprocessing system is idealized so that it causes no delays at all. It is shown that even under this idealized assumption, there are problems for which, because the parallel decompositions are inherently difficult, the optimal speed-up cannot be achieved.

Containing the cost of undergrounding

Abstract: For many years in the UK, and increasingly abroad, there has been growing pressure for the undergrounding of major overhead lines; the high cost of this has led to a vigorous development of supertension-cable technology. The paper attempts to analyse the value of the many significant advances in cable design and application in reducing the cost of cables relative to that of overhead lines. It concludes that, without these advances, the cost of supertension cables would have been some four times higher than it is. However, simultaneous development of overhead lines, coupled with changes in commercial conditions and cost of materials, have combined to halve the apparent gain from this progress, when expressed as the ratio of the cost of a cable to that of an overhead line of the same rating. In addition, operating experience shows that the reliability of supertension cables is less than that of overhead lines, to an extent that makes this an important factor in such comparisons.

Dish-stabilizers for dish washing machines

Abstract: A mesh adapted to be lowered on a group of dishes contained in a work-basket of a dish washing machine. The mesh is limp and drapes from its weight over and around the dishes to maintain them as deposited and apart from each other. A mechanized application locates the mesh overhead in the dish washing machine, and causes it to be lowered automatically as stated on the insertion of the dish-loaded work-basket into the machine and raised back to the overhead position on the withdrawal of the work-basket.

Abnormal vibration safeguard for multiconductor overhead transmission cable

Abstract: PURPOSE: Mapping out the decrease of the gallopping vibration generation chances by affixing the contact spacers with specified radius on each element conductor of the multiconductor overhead transmission cable. COPYRIGHT: (C)1977,JPO&Japio

Analysis of Deadlock Avoidance Schemes and Resource Utilization for Non-Preemptible Resources,

Abstract: : Two aspects of the performance of deadlock avoidance schemes are studied. The first is the cost of deadlock avoidance algorithms. This represents the system overhead. The second is the resource utilization under different schemes. For the first, the basic cost is the computation of the boolean function safe(I), where I is an integer vector representing the system state. safe(I) is true if the system is safe, false otherwise. The resource allocator will make the allocation only when the resulting system has safe(I)=true. Based on the concept of computation trees, several lower bounds for the cost involved in computing safe(I) are established under different conditions. An upper bound is also developed.

Numerical Performance of Matrix Inversion with Block Pivoting

Abstract: : An experiment with matrix inversion using block pivots is presented. Large scale matrix computations can often be performed more efficiently by use of partitioning. Such matrix manipulation lends itself to paged or cache memory systems since computation is staged to be completely performed in local blocks of controllable size. On other systems retrieval overhead can be balanced with computation for 'in-memory/out-of-memory' applications. Parallelism in such schema leads to efficient utilization of some multiple processor environments. Timing results indicate, however, that choice of block size should not necessarily be dictated by hardware page size for most efficient operation and that classical methods of estimating computation times are not always adequate.